Farewell to Species - reticulation

[Byron J. Adams]

Thomas DiBenedetto wrote:
 ----------------
> [snip] ... the name of the original taxon was meant to apply to all descendants of
> the common ancestor of the original taxon. So the original name applies to
> all parts of this example; the singular branch as well as both sublineages
> after the divergence. The original taxon is a higher taxon. You are right
> that there are two species, but there is also a third taxon; the higher
> taxon which encompasses the two species. And the original name must refer to
> the higher taxon.

Right on.  Except I don't see any evidence that there are two extant
species.

> Apomorphies are the evidence by which we recognize lineage divergences.

I agree.  Which is why I am cautious about your ability to recognize two
extant species when at least one of them has no autapomorphies (evidence
of lineage independence).

>A lack of evidence may make lineage identification difficult or impossible,
> but that doesnt change the underlying reality that the divergence may have
> happened. If we knew, independently of character evidence, that a divergence
> has occurred, surely we would want to recognize that,,no?
>

Yes.  But - As you point out, identifying lineages without evidence is
impossible.  The practice of recognizing species "in the absence of
evidence" (absence of autapomorphies) is essentially the substance of
the "metaspecies" concept.  I'm hoping I can convince you that
recognizing two distinct species in the ongoing example is irrational --
because it does so in the absence of character evidence.  Delimiting
species requires evidence of lineage independence.  Such evidence can
only come from independently evolved, unique, characters
(autapomorhpies).  In the ongoing example, only one lineage (the one
that budded) has evolved anything like a unique, derived character.

Assuming the taxa in question are monophyletic, and if they arose via
"descent with modification" (cladogenesis) I see a couple of problems
with this notion (recognizing two sister taxa as sister species, one
taxon maintaining the ancestral condition, the other evolving an
autapomorphy):

1.  The "species" that does not evolve an autapomorphy (I'll call it
lineage A) is at the same time an ancestor and a descendent.  This is
logically absurd.

2.  The "species" that does not evolve an autapomorphy can only be
delimited based on negative evidence (taxon A can only be defined as
"not taxon B").  Taxon A is comprised of a group of entities, the
essence of which is absence, (a privative group).  Privative groups
cannot be subdivided, and no species can be defined as the absence of
nothing.

I agree with Tom's justification for recognizing the ancestral lineage
as being distinct.  I just disagree that the descendent lineages in the
example can be delimited as species, because at least one of the two
lineages fails to exibit evidence that it is evolving independently.
The argument boils down to whether or not we think "metaspecies"
represent anything "real" (a _real_ species, as opposed to something
more than a temporally disjunct population with variation).
Recognizing metaspecies (and speciation via "budding") is guessing at
the future (as are all species delimitations), but without any evidence
of historical lineage exclusivity.

Byron Adams
University of Florida
Dept. of Entomology and Nematology
Natural Area Drive
PO Box 110620
Gainesville, FL 32611-0620
(352) 392-1901